Group 1 - Nvidia is expected to report a revenue growth of 52.4% year-over-year to $45.78 billion and a net profit increase of approximately 40% to $23.175 billion [4] - Apple has scheduled its fall product launch event for September 9, where the iPhone 17 series is anticipated to be unveiled [5] - Google has officially launched its advanced image generation and editing model, Gemini 2.5 Flash Image [6] Group 2 - Meta's AI glasses, Hypernova, are expected to enter mass production in Q3 of this year, with a price point of around $800 [7] - Okta reported Q2 revenue of $728 million, with adjusted earnings per share of $0.91, exceeding expectations and raising its full-year guidance [14] - Zhihu's Q2 revenue reached 717 million yuan, with an adjusted net profit of 91.3 million yuan, marking three consecutive quarters of Non-GAAP profitability [15] - Shangshang R&D's Q2 revenue was $139 million, a year-over-year increase of 58.7%, with a Non-GAAP net profit of $44.51 million, nearly an eightfold increase, both hitting record highs [16] Group 3 - Kroger announced layoffs affecting fewer than 1,000 employees [10] - Newmont, the world's largest gold mining company, plans significant layoffs as part of its cost-cutting strategy [11] - Rio Tinto is streamlining its product team structure into three main business segments: iron ore, aluminum and lithium, and copper [12] Group 4 - The FAA plans to conduct exercises with Boeing before considering lifting the monthly production limits on the 737 MAX [8] - Stellantis has suspended its development plan for its first Level 3 autonomous driving assistance system [9]

凤凰网科技讯 8月27日，科技媒体The Decoder昨日（8月26日）发布博文，报道称谷歌DeepMind正式推 出其最先进的图像生成与编辑模型Gemini 2.5 Flash Image，在Gemini应用中可根据文字指令，在保持人 物与动物的外观一致的情况下，提高修改图片的精度。 目前用户可通过Gemini App、API等方式访问，其API定价为每百万输出token30美元，单张图像成本约 0.039美元。 据介绍，与此前的原生图像生成工具相比，它在根据文字进行图像修改时的准确率更高，谷歌表示， Gemini 2.5 Flash Image在多项任务中优于ChatGPT所用的GPT-4o，特别是在编辑图像时，它可以根据文 字提示进行编辑。 该模型还支持通过文本提示进行精确的局部编辑。用户可以通过模糊背景、去除瑕疵、添加颜色或擦除 整个物体，无需进行手动选择。同时该模型支持一次融合最多三张图像。 ...